Costing us slowlypart one here.]
After the fixed costs, then you've got to deal with the vagaries of incremental cost. Creating an e-book isn't free, and we are talking about running a business. (You want your authors to eat something other than ramen, don't you?)
Let's pretend you create a book. The fixed costs of producing the book at all was $100. Each paper book will cost an additional $1 to ship and converting it to an e-book will cost $5.  That is, if you make 100 paper copies it will cost you $100, and if you make 100 e-book copies, it costs you $5. That's an argument for cheaper e-books, right?
If you sell one e-book, the cost-per-unit is $5.
If you sell one paper book, the cost-per-unit is $1.
The cost-per-unit price (or incremental cost) for each additional e-book goes down as you sell more copies. With our overly-simplistic example, you'd have to sell five e-books and five paper books for the incremental costs to be the same. Poor sellers will have a higher cost-per-unit than good sellers when you talk about e-books . That's much harder to plan for.
"But Steve," my hypothetical person says, "you've got to do most of those sunk costs anyway for a print book. Yeah, there's some wonkiness about layout formats that needs tweaking, but that can change. You've already sunk that money into paper book layouts, and conversion costs aren't that high. Are the publishers greedy? Why not pass on the savings?"
Those fixed costs don't go away. Each book and e-book sold chips away at the fixed costs. (e-books will do so at a variable rate, as illustrated above.) The argument my hypothetical person made is based on having paper books subsidize the fixed costs for e-books.
Right now the e-book market is less than 5% of total book sales. Penguin says that e-books are 4% of sales now and hopes to make them 10% of sales by next year. I think that prediction is pretty spot-on; I've seen several previously anti-e-book people become converted suddenly in the last few months. Additionally, if we were to price e-books lower than regular books, I'd guess that would accelerate that conversion from paper to e-book use.
The more people move to e-books, the more of the sunk cost will be borne by e-books. That totally blows my hypothetical person's argument out of the water.
There is more price elasticity with e-books – that is, the price can more easily go down when they're less in demand and go up when they are in demand. I think that kind of dynamic pricing is wonderful – and it's one of the reasons I think Amazon was on the wrong side of the argument with McMillian a little while ago. Price elasticity is where you will see e-books be (sometimes) cheaper than paper books.
 An absurdly low (and pulled out of thin air) number for the sake of illustration. Also, I'm aware that publishing is MUCH more complicated and bizarre than this.
 This is actually true of paper books as well.